Triple
T38084012
| Position | Surface form | Disambiguated ID | Type / Status |
|---|---|---|---|
| Subject | Léogâne commune |
E950925
|
entity |
| Predicate | reconstructionEffortsBy |
P86095
|
FINISHED |
| Object | Haitian government |
—
|
NE NERFINISHED |
How this triple was built (2 steps)
Every LLM step that produced this triple, in pipeline order — named-entity classification, the disambiguation choices (the exact options shown, with the pick highlighted), and the generated description. The batch + timestamp of each is in the Provenance table below.
NER
Named-entity recognition
gpt-5-mini
Instruction
Given a phrase, classify it is english named entity (e.g., persons, organizations, works of art) in Latin script, or not (e.g., literals, dates, URLs, verbose phrases). For disambiguation, the statement where the phrase occurs as object is also given. Please return a JSON object with `phrase` (string, the phrase being analyzed) and `is_ne` (boolean, indicating whether the phrase is a Named Entity).
Input
Phrase: Haitian government | Statement: [Léogâne commune, reconstructionEffortsBy, Haitian government]
PD
Predicate disambiguation
gpt-5-mini-2025-08-07
Target predicate: reconstructionEffortsBy Context triple: [Léogâne commune, reconstructionEffortsBy, Haitian government]
-
A.
reconstructionWork
Indicates that an entity is engaged in or associated with activities to rebuild, restore, or repair something that was damaged, destroyed, or altered.
-
B.
reconstructionWorkBy
chosen
Indicates that a reconstruction project or effort is carried out, overseen, or authored by a specific agent or entity.
-
C.
reconstructionIncorporated
Indicates that one reconstruction has been incorporated into, or included as part of, another reconstruction.
-
D.
revitalizationEffort
Indicates efforts or actions undertaken to restore, renew, or improve the condition, vitality, or functionality of something.
-
E.
hasReliefEffortFrom
Indicates that a party receives assistance or support through a relief effort provided by another party.
- F. None of above.
Provenance (3 batches)
The batch behind each pipeline step, in order, with when it ran. Timestamps are batch-level — stages were processed in waves, so the object chain (NER → NED1 → NEDg → NED2) reads in order, but predicate / elicitation batches can sit in a different wave.
| Step | Stage | Batch ID | Status | When |
|---|---|---|---|---|
| creating | Elicitation | batch_69f76f03a3608190a73fd6df87c792a8 |
completed | May 3, 2026, 3:51 p.m. |
| NER | Named-entity recognition | batch_69fc4748843c8190931432653be4890c |
completed | May 7, 2026, 8:03 a.m. |
| PD | Predicate disambiguation | batch_69fc45646ce481908caf292ff9f06e15 |
completed | May 7, 2026, 7:55 a.m. |
Created at: May 3, 2026, 4:21 p.m.